Work Engagement Scale: Construct Validity and Reliability in the Colombian Organizational Context

Int J Psychol Res (Medellin). 2023 Jun 20;16(1):114-125. doi: 10.21500/20112084.6408. eCollection 2023 Jan-Jun.

Abstract

This manuscript shows the results of the evidence of content and internal structure obtained from an instrument to measure work engagement. This instrument is aimed at workers of Colombian labor organizations that belong to different economic sectors (commerce, services, education, and health). The instrument was designed based on the postulates of cognitive theory and is structured into three factors that operationalize the construct: (a) behavioral dimension (15 items), (b) affective dimension (16 items), and (c) cognitive dimension (14 items), for a total of 45 items. The results of the content evidence through expert judgment suggested the elimination of three items, due to ambiguity and lack of clarity, leaving the 42item test. After this evidence, the instrument was piloted in a sample of 460 participants. The item-test correlation analysis recommended the elimination of one item due to its low correlation with the factor. The evidence of internal structure through exploratory factor analysis (EFA) proposed a three-factor structure, with an explained variance of 63%; 9 items were eliminated due to high residual. The fit indicators showed a GFI = .99; and those of residual showed a RMSR =.03 and Kelley = .04; each factor obtained an ordinal Cronbach's Alpha of .95 (behavioral), .97 (affective), and .87 (cognitive). These results indicate precision in the measurement and consistency of the items to measure each of the factors.

El presente manuscrito muestra los resultados de la evidencia de contenido y estructura interna obtenidas de un instrumento para medir work engagement. Este instrumento está dirigido a trabajadores de organizaciones laborales colombianas que pertenecen a diferentes sectores económicos (comercio, servicios, educación y salud). El instrumento se diseñó basado en los postulados de la teoría cognitiva y se estructura en tres factores que operacionalizan el constructo: (a) dimensión conductual (15 ítems), (b) dimensión afectiva (16 ítems), y (c) dimensión cognitiva (14 ítems), para un total de 45 ítems.Los resultados de la evidencia de contenido a través del juicio de expertos sugirieron la eliminación de tres ítems, por presentar ambigüedad y poca claridad, quedando la prueba de 42 ítems. Posterior a esta evidencia, se piloteó el instrumento en una muestra de 460 participantes. El análisis de correlación ítem-test recomendó la eliminación de un ítem por presentar baja correlación con el factor. La evidencia de estructura interna a través del análisis factorial exploratorio (AFE) propuso una estructura trifactorial, con una varianza explicada del 63%; se eliminaron 9 ítems por presentar alto residual. Los indicadores de ajuste mostraron un GFI = .99; y los de residual un RMSR = .03 y Kelley = .04; cada factor obtuvo un Alfa de Cronbach ordinal de .95 (conductual), .97 (emocional) y .87 (cognitivo). Estos resultados indican precisión en la medida y consistencia de los ítems para medir cada uno de los factores.

Keywords: Engagement; Engagement Scale; Instrument Validation..